AI-Assisted Adaptive Control for Operation Optimization of Smart Inverters

Abstract: Inverter, as a power electronic device that realizes DC-AC conversion, plays an indispensable role in many industrial fields such as renewable energy grid-connection, electric vehicle traction, motor drive and so on. How to design an efficient and reliable control strategy has been the goal of researchers for a long time. However, in real working conditions, there may be lack of sensors, inaccurate circuit parameters and other adverse conditions, so it is of practical significance to design a controller for such control problems caused by information loss.
The focus of this research project is to exploit the potential of artificial intelligence for feature recognition to optimize the operation of smart inverters while maintaining the high accuracy of classical model-based control methods. First, it is planned to propose an AI-assisted adaptive control framework, and then apply the proposed control framework to actual industrial scenarios to optimize the existing control effects of EV traction inverters and PV grid-connected inverters. The expected results will provide ideas for optimising the operation of related inverter applications.

Funding: CSC Scholarship